Through the onboard monitoring test, sensor fault in IKOMASAN and unstable measurement situation in NYK ANTARES were experienced, and it was found the former was due to the damage of the zirconia by moisture adhesion in the washing of exhaust gas economizer and the latter was due to the dust adhesion by reduced temperature inside the exhaust gas sampling line. On these problems, the countermeasures have already been established as above-mentioned and it is in the situation that the effectiveness has been confirmed. Also on the accuracy of the zirconia sensor, the drift of the zero/span is slight, and the accuracy without the problem in the practical use has been obtained. Regarding the NOx monitoring device, including these matters, the concrete correspondence in maintenance aspect and necessary frequency of calibration is reported in the clause 6.4, and here, the details are omitted.

　

 Through the test by this fiscal year, the problems in the application to the in-service ship as a NOx monitoring device have been clarified, and the countermeasures on those have been also established. Therefore, it was confirmed that the continuous measurement with sufficient reliability and accuracy was possible, if the calibration and the replacement of parts were carried out at adequate timing.

　

 NOx emission rate on shipboard of IKOMASAN can be found to be within a range from - 10% to +15% relative to shop trial data.

 In case of NYK ANTARES, NOx emission rate on shipboard is distributed within a range from -10% to +20% relative to shop trial data.

　

 NOx emission rate on shipboard of B ship is distributed within a range from +5% to +30% relative to shop trial data, though the number of data is still little. In case of B ship, the measurements have been made using this monitoring device since her sea trial. Although NOx emission rate at sea trial by using distillated fuel oil is equivalent with that of shop trial, NOx emission rate at sea trial by using residual fuel oil is distributed within a range from +15% to +25% relative to shop trial data. This shows clearly that the difference of generated NOx has occurred only by the difference of fuel oil.

　

 By the results of these ships, it was confirmed NOx emission rate on shipboard using residual fuel oil could be measured in the level over +15% relative to shop trial data, though the dispersion by various factors can be considered. In the future, fluctuation factors concerning the combustion which originates from the fuel property, etc. will be examined, and the current onboard allowance (=15%) must be insufficient.

　

 Among these 3 ships. the range of the relative level of NOx emission rate to shop trial data is slightly different one another. Though it can be thought to be caused by a difference in combustion which originates from engine types and used fuel, etc., the effect by data on the intake air humidity (temperature and relative humidity of the intake air ) which is one of engine operating condition data is also considered. (The main factor which affects humidity correction factor KHDIES is the absolute humidity of intake air)

　

 In this test, the electric digital thermohygrometer is used in IKOMASAN, and the psychrometers are used in other 2 ships. The effect by the difference in measurement procedure has not been examined, and the degree of the correction of intake air condition will be evaluated using data measured by the same condition, and the factor of the difference in the level of NOx emission rate will be examined, in future test.

　

 Regarding the fuel oil used on the objective ship, most of samples have a content of nitrogen within a range from 0.3% to 0.4%, and the result of NOx conversion rate of nitrogen content by constant volume combustion test using them was obtained as 55%.

　

 Though it depends on fuel oil consumption, NOx emission rate by the fuel of 0.4% nitrogen content comes up to nearly +10% ( due to the increment of Fuel NOx) compared with that by distillated fuel oil. And it was proven that generated NOx (Fuel NOx + Thermal NOx) might exceed +15% only by the difference in the fuel oil, considering the fluctuation of the Thermal NOx by the difference in combustion (heat release rate pattern). Although this is not a result of the combustion test on the actual engine but a referential information, it shows sufficiently that NOx emission rate over +15% relative to shop trial data can be measured on shipboard only by the difference of the fuel oil.

　

 Further verification on the NOx conversion rate, etc. is necessary, and the engine test on a single cylinder test engine has been scheduled. In future, the evaluation on generated NOx with the fuel oil property (nitrogen content and combustibility) will be advanced.

　

 In this test, the relationship between engine load and NOx emission rate on shipboard was evaluated based on the relationship between engine load and NOx emission rate at shop trial. By quantifying the NOx emission rate in a relative proportion to shop trial data, NOx emission rate on shipboard has been evaluated regardless of the E3 cycle value, and it is regarded as an effective evaluation method in the meaning of the relative comparison with shop trial. It seems to be the most realistic to evaluate the onboard monitored data at engine load of 1 point (normal load) or 2 points in this method.

　

 Future investigation on the Allowance, which appropriately reflects the effect of generated NOx originated from the fuel oil and the measurement error onboard, will be necessary, promoting the accumulation of measured data on shipboard in various conditions.

　

 Based on information got by this onboard NOx monitoring test, NOx monitoring method can be effectively utilized by following:

　

 As mentioned above, nitrogen contained in fuel oil has big influence on NOx emission. Fig 8.3.1 shows the relation between nitrogen content in fuel oil and NOx emission measured at the second test ship. There is some dispersion also, however, tendency that NOx emission proportionally increase by nitrogen content could be seen. For an example, when calculate by using the linear approximation line in the figure, the NOx increment of 12% could be got at 0.4% nitrogen fuel compare to 0% nitrogen fuel. This value shows close value with the result of the third test ship.